In a computer-implemented remote conference meeting, participants are not identified automatically and considerable amount of time is spent on each participant identifying oneself. As the number of participants increase, this task becomes even more cumbersome and prone to confusion. Due to the globalization of workforce and current trends in individuals working from remote offices, for example, it is expected that this problem will contribute to increase inefficiency in resources. Even with authenticated devices, the problem remains, since multiple participants may be located in or associated with the same device. Therefore, there is a need to automate the task of automatically identifying the speakers in the conversation, for example, in online meetings or chat rooms, in a pervasive way such that the cadence of the meetings are not changed due to the necessary speaker identification.